Debt collection goes hi-tech
According to the National Credit Regulator, of SA’s 24-million credit active consumers, some 10-million have impaired credit records. “We hold data on 9.2million of these nonperforming individuals,” says David Hurwitz, CEO of Transaction Capital, the holding company of customer credit management solutions provider Transaction Capital Risk Services.
It’s a significant statistic because the company uses that data to collect nonperforming debt from consumers as an agent on behalf of lenders, banks, insurers, corporates and SOEs, or as a principal after acquiring the debt.
“To do so we operate a bespoke fintech platform that offers an effective means to track and engage with debtors in an individualised and contextually relevant manner,” says Hurwitz. “We first run portfolios through our database to update existing data or add new information. Our data analytics engine then tracks an individual’s debt across all creditors and lenders, and the number of instances that person is flagged in the system. This information is then analysed to determine a ‘propensity to repay score’.”
The platform uses machine learning algorithms to generate an interaction model that predicts the consumer’s “contactability” and “transactability” — how consumers prefer to be contacted and settle debts. By marrying these factors, the platform tailors and individualises each collection interaction by determining the most appropriate communication channel to use, be it a call, SMS or e-mail; the language, phraseology, tone and content of the message; and the ideal method of payment.
“Any form of digital communication can then be automated by the system and tracked. For more difficult consumers, repeat offenders or large unrecoverable amounts, the system may suggest a legal letter,” says Hurwitz.
However, the majority of engagements happen through the call centre. “Here technology aids the process. We operate call centres with auto-diallers for efficiency. The system determines the appropriate number to call and what time of day or month is best.”
Transaction Capital has also implemented layered voice analytics to track and record calls to ensure customers are being treated fairly. “This artificial intelligence-enabled system can analyse a client’s voice to predict how likely they are to pay, and if they’re being honest in the answers they provide,” adds Hurwitz.
With every engagement the system refines subsequent interactions and provides updated insights and data.